In the past few weeks we have tested two CPUs in over 30 games: AMD's Ryzen 5 2600 and Intel's Core i5-8400. Before we get to the benchmark results – and I promise that there will be a shipload of them – here are a few brief tips on the test setup.
The Core i5-8400 rig has the MSI B360 Gaming Plus, a high-quality B360 board that can easily make optimal use of the hexa-core CPU. Of course, we're limited to DDR4-2666 on this card, but we chose G.Skill's premium FlareX CL14 memory with very low latency. The cooling of the i5-8400 is the standard unit from Intel.
Then we used the Asus ROG Crosshair VII Hero for the Ryzen 5 2600 and we have two configurations: a standard configuration with the Wraith Stealth box cooler and the G.Skill FlareX CL14 memory clocked at 2933 MHz. And a second test setup that runs on a 4.2 GHz all-core overclocking using aggressively tuned G.Skill Sniper X DDR4-3400 memory with tighter sub-timings. Here the cooler was upgraded to the Corsair H115i Pro.
We have 36 games on the menu for testing. Every game was tested with the GeForce GTX 1080 Ti with a resolution of 720p, 1080p and 1440p. I would have liked to take the Vega 64 Liquid too, but since we had already run 324 individual tests, three times each … almost 1,000 benchmark runs later, we weren't particularly interested in reaching these 2,000. At the moment we only focus on the CPU comparison and let the 1080 Ti work in the background. So have a drink, some snacks and make yourself comfortable. Let's go into that.
ARK: Survival Evolved, ARMA 3, Ashes of Singularity, Assassins Creed, Battlefield 1
First of all we have ARK: Survival Evolved and this is not a title that I test often, but by popular request it has been added to the list of tested games. It's often listed in the top 10 most played games on Steam, so I imagine some of you want to see the results.
I expected this to go bad for Ryzen, but the results are very surprising. We are mostly tied to the GPU, although Ryzen has an advantage over us at 720p. However, due to the GPU limits, the overclocked Ryzen 5 2600 was not faster than the standard configuration.
Another game that is very popular on Steam and is always in high demand is ARMA 3. In the past, this game ran horribly on AMD processors, including Ryzen, but I was told that the games were recently updated to include Making better use of Ryzen CPUs, or at least CPUs with more than one core, and that certainly seems to be the case.
After unpacking, the Ryzen 5 2600 had no problems with the Core i5-8400 and the games were clearly CPU limited as the overclocked 2600 offered impressive gains.
Even at 1440p we noticed a massive 20% increase in performance when overclocking Ryzen cores and memory. This meant that when comparing the average frame rate, the 2600 was now 19% faster than the Core i5-8400. I'm pretty shocked by these results given Ryzen's poor performance in the last ARMA 3 test.
Come on, you knew Ashes of the Benchmark would be included, and it's the only DX12 game that I've tested with DX12 with a GeForce graphics card. Here we see that the Core i5-8400 and the Ryzen 5 2600 are immediately very evenly matched. When testing with 1080p and 1440p, the primary GPU is bound to the activated "crazy" preset, so overclocking from Ryzen is not particularly advantageous. At 1440p there is only an increase of 5%, then 7% at 1080p and then a somewhat impressive 11% at 720p.
Okay, some interesting results in Assassins Creed Origins. Let's talk about what's going on here. You will find that the standard Ryzen 5 2600 at 1080p offers exactly the same performance as at 720p. This is because the CPU is limited. Overclocking solves this problem and at 1080p we see 2600 able to match the 8400, but both CPUs are now GPU limited. We know this because the overclocked 2600 at 720p is able to significantly prefer the 8400 for the average frame rate.
At 1440p, we're finally almost entirely tied to the GPU, although the standard Ryzen 5 2600 configuration still has difficulty maximizing the GTX 1080 Ti.
Testing with Battlefield 1 shows how good the Core i5-8400 is. He immediately delivers a very strong performance in this demanding title. The Ryzen 5 2600 is also very handsome and while it was only 6% slower for the frame time result at 1080p, the average frame rate was 12% lower. At 1440p, this latitude is somewhat removed, since we are strongly tied to the GPU. It is interesting to note, however, that the 2600 gives a much better 1% result even at this resolution.
At the lower resolutions, we see that overclocking the 2600 improves the average frame rate by almost 20%, and while this is an impressive gain, it makes the tuned Ryzen processor just a few percent faster than the Core i5-8400. But here, too, the 1440p results are the most impressive. Here the 2600 was 18% faster than the 8400 for the frame time result.